Why GraphQL vs. Why not GraphQL
summary
Claims that GraphQL removes REST overfetching by letting the client request fields, provides schema-as-documentation, and uses a single endpoint; shows a working prototype built in Java and notes practical tradeoffs. Demonstrates problems: client-controlled queries can spike backend cost so use persistent queries or depth limits, federation and naming congestion, ORM impedance, caching loss versus CDN, and awkward numeric types like 32-bit integers requiring custom scalars.
AI-written summary · about
chapters (18)
- 0:00Introduction
- 0:48Disclaimer
- 1:40What is GraphQL
- 3:22Limitations of REST
- 3:56GraphQL vs REST
- 4:22Wardrobe example
- 5:32GraphQL
- 5:58GraphQL Schema
- 6:43Request Data
- 8:30Common Use Cases
- 10:31GraphQL is new
- 12:48Carbo cultic
- 15:16Control
- 17:59Centralization
- 22:44Tech Stack
- 24:37Performance
- 25:43Over fetching
- 28:22Supriya introduction
